Which methods do you think would be most effective in consolidating the results so that the assumptions drawn from the sampdfple data will better reflect the population?